Computer Engineering and Applications ›› 2009, Vol. 45 ›› Issue (10): 199-201.DOI: 10.3778/j.issn.1002-8331.2009.10.060

• 工程与应用 • Previous Articles     Next Articles

Research on expert system for mould level control

LI Zu-lin   

  1. Department of Electric and Information Engineering,Hunan Institute of Engineering,Xiangtan,Hunan 421002,China
  • Received:2008-05-05 Revised:2008-09-08 Online:2009-04-01 Published:2009-04-01
  • Contact: LI Zu-lin

结晶器液位专家系统控制研究

李祖林   

  1. 湖南工学院 电气与信息工程系,湖南 湘潭 421002
  • 通讯作者: 李祖林

Abstract: Level control project has disadvantage for conventional mould control system.According to mechanism model,compare mould level control effect based on Jam-stick Proportion-Integral(PI) and casting speed PI conditioner and analyze question of Jam-stick PI conditioner,controller based on Redial Basis Function Neural Networks(RBFNN) and casting speed compensator are designed.Coordination control system is studied based on expert system of Jam-stick and casting speed regulating by using rule model.Interfere is resolved for switching level regulator.Expert system is simulated based on MATLAB,and tested in practice.The results show that level control accuracy and stability and dynamic response excel conventional single-loop control project.

Key words: expert system, neural networks, mould, level control

摘要: 针对传统的结晶器液位控制系统方案存在缺点,根据对象机理,比较了塞棒比例积分(PI)控制和拉速PI控制和对结晶器液位控制效果,分析了塞棒PI控制液位存在的问题,设计了基于径向基函数神经网络(RBFNN)的塞棒控制器和拉速补偿器。使用规则模型,研究了基于专家系统的塞棒和拉速控制液位的协调控制系统,解决了液位控制器切换存在的干扰。系统在MATLAB上进行了仿真,并在实际中进行了测试。实际测试结果表明该系统的液位控制精度、稳定性和动态响应优于传统的单回路控制方案。

关键词: 专家系统, 神经网络, 结晶器, 液位控制